Go Back
Dynamics of Organizational Change
Özge Pala
Jac Vennix
Nijmegen School of Management
Nijmegen University
P.O. Box 9104
6500 HE Nijmegen
phone: +31-24-3615568/fax: +31-24-3612351
email:O.Pala@maw.kun.nl
Keywords: Strategic change, punctuated equilibrium, strategic fit,
performance
appropriateness of strategy,
Abstract
This paper discusses a model of organizational change as described in the punctuated equilibrium
model. The model builds on two previous models in this area, i.e. the one by Frechette and Spital and
the one by Sastry. After discussing shortcomings in both these models a new model is presented which
is believed to be (a) a more valid representation of Tushman and Romanelli’s theory of punctuated
organizational change, and (b) an extension of both existing models incorporating new structure not
represented so far. The paper discusses the structure of the new model and tests its validity by
comparing dynamics from the model with those stated in the theory. Sensitivity analyses are
conducted in order to further explore model behavior and find potential inconsistencies and areas for
improvement. In a concluding section the limitations of our model are discussed and improvements for
the future are proposed.
1. Introduction
In the last decades, many theories have tried to explain the reasons and consequences of
organization change. The most prominent of these are the population ecology, adaptation theories, and
the punctuated equilibrium model of change. The last one of these is the most detailed one with
respect to the causal relationships. It is also a very widely accepted theory. For example, Gersick
(1991) tries to show the high applicability of this model for organizational studies by discussing
models from six different domains such as groups, history of science, and biological evolution. In the
SD literature, Frechette and Spital (1991) and Sastry (1997) have used the punctuated equilibrium
model by Tushman and Romanelli (1985) to construct dynamic models of organizational change.
Tushman and Romanelli’s punctuated equilibrium model of change emphasizes two types of
change: incremental and strategic. During an incremental change period, also called the convergent
period, the organization tries to “achieve a greater consistency of internal activities with the strategic
orientation”. The strategic changes, reorientations, are defined by “simultaneous and discontinuous
shifts in the strategic orientation”. According to Tushman and Romanelli, “patterns of organizational
evolution are characterized by periods of convergence punctuated by reorientations leading to the next
convergent period” or “organizations evolve through relatively long convergent periods punctuated by
discontinuous or metamorphic changes which lead to the next convergent period”.
Even though Tushman and Romanelli make dynamic propositions on what characterizes both
incremental and strategic change, in this paper, we’ll focus on strategic change because it is the
dynamics of the radical change process we are mainly interested in. Incremental change is important in
the sense that what happens during the convergent period affects the trigger mechanism behind and
success of strategic change. The effect of a convergent period will be represented with exogenous
variables for the time being. Future versions of this model will make the convergent period
endogenous.
1
In the next section we will discuss the two previous system dynamics models of organizational
change and indicate their most important shortcomings. After spotting these deficiencies, a new model
of the punctuated equilibrium theory will be proposed. Next the model will be tested using extensive
sensitivity analysis from which a number of conclusions will be derived. The final section of this
paper contains a critical discussion of our model, relates conclusions to existing literature and
proposes required changes for the future.
2. SD models of punctuated equilibrium model
As stated, there are two SD models of the punctuated equilibrium model. One is by Frechette
and Spital (1991) and the other is by Sastry (1997). First, the model of Frechette and Spital (1992) will
be explained since it precedes the latter in time.
2.1 The model of Frechette and Spital
Frechette and Spital’s (1991) objective was “to model the key components of a change system;
to describe the relationships between components and to explore policy options…to improve the long
run effectiveness of organizational change” (p209) by taking the theory explained in the 1985 paper of
Tushman and Romanelli as a basis. They see two important dimensions of change: scope of change
and timing of change. Scope of change signifies the difference between incremental and strategic
change processes, both of which are modeled explicitly. Timing of change refers to reactive vs.
anticipatory change processes. Only the prior one, i.e. change initiated as a response to an event (or
series of events) that has occurred, is modeled.
In the model (see figure 2), change is initiated only after ‘readiness to change’ reaches a
certain level. ‘Readiness to change’ is affected both by perceived performance gap and by the
organization’s memory with respect to success or failure of previous change experiments. After
performing policy experiments to understand the relative importance of perceived performance gap vs.
organization’s memory on initiating change, Frechette and Spital conclude that performance
improvements are sustained in the long run only when there’s a balance between the effects of
performance and organizational memory (p217).
start incremental change
resources for incremental
failed incremental improvements
readiness for incremental change
successful incremental improvements
perceived performance gap
successful strategic improvements
readiness for strategic change
failed strategic improvements
resources for strategic
start strategic change
Figure 1: The feedback structure of the model by Frechette and Spital (1991)
The study of Frechette and Spital is an important one because it takes as a basis an accepted theory of
organizational change. The explicit distinction between incremental and strategic change is an
important step. The simulation experiments shed light on the existing theory of punctuated equilibrium
model as well. Even though Tushman and Romanelli mention the effects of performance gap and
success/failure of previous changes on inertia (‘readiness for change’ in the model can be interpreted
as inertia), they do not discuss the relative importance of the two. The simulation experiments by
2
Frechette and Spital show that if one of these is given higher priority than the other for initiating
change, change leads to performance crisis.
Problems with the model of Frechette and Spital
The model of Frechette and Spital and the policy experiments they run, show the possibility of
modeling the issues around organizational change and using this model for answering various policy
questions. Yet the model is lacking some important factors that influence the dynamics of a change
process within organizations.
The explicit modeling of the concept of inertia should be included. One may argue that
“readiness to change” represents inertia. However, there is a problem with this argument such that in
the model, readiness to change applies both to incremental and strategic change whereas inertia only
applies to strategic change. Tushman and Romanelli describe inertia as “resistance to all but
incremental change” (p. 177). Thus, inertia does not form a constraint for incremental change, but
rather it builds up during the period of incremental change.
One of the drivers of change is the performance gap. Performance gap is modeled more or less
as an exogenous variable. The only thing that affects performance is the success/failure of change.
However, in order to trigger change, performance is given a pulse, rather than having performance get
calculated by its determinants. Tushman and Romanelli are very clear about the determinants of
performance: internal and external consistency of strategic orientation. We believe that it is important
to make performance totally endogenous by explicitly modeling its determinants. Moreover, the only
determinant they modeled, i.e. the success/failure of change, does not have a direct effect on the
performance gap. As change takes place, the fit of the strategic orientation to the environment gets
affected. Consequently, the level of the performance also changes. The explicit modeling of these
relationships would yield a more structurally valid model that represents the theory more precisely.
In relation to the previous point, it should also be pointed out that the external stimulus for
change is created by giving a pulse to the performance gap. According to Tushman and Romanelli,
when a change occurs in the environment, it increases the probability of strategic change for the
organization. Thus rather than giving a pulse to performance gap, a change should be created in the
environment which leads to change in the external fit of the strategic orientation. This would lead to a
decrease in the performance or increase in the performance gap, which might lead to strategic change.
It would then also be possible to see the effects of the response time of the organization, i.e. the time
between a change in environment and strategic change, on the long-run performance of the
organization.
2.2 The model of Sastry
A significant contribution to modeling the punctuated equilibrium model is the model by
Sastry (1997). Her aim was “to examine the completeness, consistency, and parsimony of the causal
relationships laid out” (p237) in Tushman and Romanelli’s 1985 paper.
The model (see figures 3 and 4) includes only strategic change and consists of four sectors:
strategic orientation, inertia, performance, and pressure to change. Change is initiated as a result of a
performance gap, i.e. the difference between the desired performance and the perceived performance.
Performance has two drivers: appropriateness and competence. Appropriateness shows whether the
strategic orientation of the organization is in alignment with the strategic orientation required.
Competence shows the degree of efficiency the organization has in fulfilling its tasks. Competence is
determined with the inertia. When the performance gap is high, a change in strategic orientation takes
place provided that the organization has enough ability to change. This change has two effects. First, it
makes strategic orientation come closer to the required strategic orientation. Second, it decreases the
inertia.
As a result of experiments, Sastry found some gaps in Tushman and Romanelli’s theory. One
of the gaps she spotted was inappropriate, ongoing reorientation leading to organizational demise:
“Once change was initiated, organizations continued to reorient inappropriately beyond the optimal
point, eventually resulting in organizational collapse” (p256). To solve this problem, Sastry added a
measure of organization-environment fit. Now, the decision makers do not decide to go through
change when there is a pressure to change but they use their perceptions of organization-environment
fit in deciding whether to change and how much to change the organization’s strategic orientation in
3
response to given levels of pressure to change. This way the perception of a mismatch between the
organization and the environment along with sufficient ability to change and high enough pressure to
change affect the organization’s change process.
- a p p r o p ria ten ess
+
stra teg ic or ien ta tio n r e q u ir ed
+ p e r fo r m a n c e
stra te g ic or ien ta tio n
+
p e r c eiv ed p er fo r m an ce
c h a n g e in stra teg ic or ien ta tion
+
p r e ssur e to ch an g e
+co m p eten ce
+
in e rtia
++
lea rn in g , so cia liz a tio n , etc .
a b ility to c h a n g e
Figure 3: The feedback loop structure of Sastry’s model (1997)
inertia_decrease
inertia
inertia_increase
fractional_inertia_decrease
competence
fractional_inertia_increase
SO_shortfall
perc_SO_shortfall
strategic_orientation
ability_to_change_SO
SO_change
approprateness
fract_ch_press_decr
SO_indicator
SO_shortfall_indicator
change_press_incr
perc_SO_change
pressure_to_change
change_press_decr
appr_times_comp
change_indicator
trial_period_routine
perception_time
performance_shortfall
performance
trial_indicator
perceived_performance
trial_indic_infl
trial_indic_outfl
perc_perf_change
Figure 4: The stock-flow structure of Sastry’s model (1997).
The other improvement Sastry added was a trial-period routine: “This routine enforces a
waiting period following each reorientation during which executive leadership refrains from making
further change”. This routine was required because Sastry realized that during the simulation,
organizations got engaged in change efforts too soon right after a previous reorientation and thus the
organization did not have any opportunity to rebuild the destroyed competence. With this routine, “by
suspending the change process after a given level of organizational change has been undertaken,
organizational leaders using a trial-period routine protect the organization from collapse”.
The study of Sastry has important implications in terms of showing the usefulness of
simulation for analyzing the dynamic consequences of a theory in a field that is primarily dominated
4
by verbal theories. However, we believe that the model has some deficiencies, which will be discussed
in the next section.
Problems with the model of Sastry
In Sastry’s model, performance is determined by competence and appropriateness. Whereas
the appropriateness is correctly determined by the organization-environment fit of the strategic
orientation, the cause of competence is not correctly represented. In her model, competence is caused
by inertia, “inertia competence”. However, this is not a correct representation since competence is
built as a result of processes that also lead to the generation of inertia. According to the theory, in time
“interdependent relationships with buyers, suppliers strengthen, commitment to internal participants
and external evaluating agents are elaborated into institutionalized patterns of culture, norms, and
ideologies”. In turn, these “emergent social and structural processes” have two outcomes: (1) inertia,
(2) convergence on a strategic orientation through enforcement of rules and norms, which lead to the
development of a high degree of competence in executing a strategic orientation. Then, the
relationships should be represented as “social and structural processes inertia” and “social and
structural processes convergence on a strategic orientation competence” rather than “inertia competence”.
If our understanding is correct then it renders some of Sastry’s conclusions ineffective in terms
of assessing the theory’s capabilities of explaining organizational change. In one of the simulations,
Sastry realized that an organization with low inertia (representing a young organization or an
organization right after change) gets into a process of “inappropriate and ongoing reorientation leading
to organizational demise”. The reason is as follows: even though the organization has maximum value
for appropriateness, low inertia leads to low competence which is sufficient to generate a low enough
performance that causes high pressure to change and eventually the organization reorients. Since every
reorientation leads to decrease in inertia, the organization gets trapped in a vicious circle of
unnecessary reorientations. However, if the performance is modeled to be a result of internal and
external consistencies (neither of which are caused by inertia) then there is still the possibility that an
organization with low inertia will not change given that its strategic orientation is both internally and
externally consistent.
There is also a problem with the representation of strategic orientation. It is modeled as a
unidimensional state variable, because “it is the outcome of the accumulation of past decisions on
strategic orientation”. This variable is modeled as “a continuous expression that represents
management’s policy of response to pressures to changes” (p250). Then the consistency of the
strategic orientation is calculated by the difference between the required and actual strategic
orientations. We believe that this way of quantification of strategic orientation is not correct since
strategic orientation is not a continuous variable. Let’s give a tangible example. Technology is part of
strategic orientation. The firm’s technology can be, for example, A, B, or C. Let’s say it is A and the
required strategic orientation is B. What does that mean? How can we quantify these two things and
subtract them from each other to find the gap. We believe that what should be operationalized is not
the strategic orientation itself but the fit of the strategic orientation. As will be explained later, this is
the important aspect of strategic orientation for organizational change.
In the base model, the SO indicator is 1, meaning that the organization does not measure
organization-environment fit at all. Simulations with this formulation result in an organization that
changes inappropriately and continuously. As a result, Sastry concludes that there is a gap in the
theory. She states “ I found that decision makers must also measure and respond to organizationenvironment fit”. Thus, she adds a routine that also calculates the organization-environment fit and in
the second case, the strategic orientation change is affected not only by high enough pressure to
change, but also a perception of mismatch between the organization and the environment. While this
finding is true, it already existed in Tushman and Romanelli’s theory. In the theory, even though
Tushman and Romanelli say that performance pressures are very important in causing change, they
also assert that changes in the environment are very important forces for reorientation (p178). And the
organizations change in order to achieve a better fit between the existing strategic orientation and the
strategic orientation dictated by the environment. Thus, we believe that this particular routine is
already a part of Tushman and Romanelli’s theory and the simulations without this routine do not
show a gap in Tushman and Romanelli’s theory.
5
After having discussed the models constructed by Frechette and Spital and by Sastry we will
now turn our attention to our model of organizational change.
3. The new model
The model consists of 3 main parts: (i) the strategic orientation, (ii) the felt pressure to change,
its determinants and completed changes, and (iii) appropriateness of the new strategy.
3.1 Strategic orientation
Strategic orientation is defined in terms of “the business the firm is in and how it competes”. It
can be described by the set of organizational activities consisting of (1) core beliefs and values, (2)
products, markets, technology, and competitive timing, (3) the distribution of power, (4) the
organization’s structure, and (5) the nature, type, and pervasiveness of control systems.
Even though, the definition and understanding of strategic orientation is important, what is
significant in terms of dynamics is not what the strategic orientation of the firm is but rather its
meaning for or its effect on the strategic change process. Although Tushman and Romanelli explain
what strategic orientation is, their main focus is on the appropriateness, effectiveness, alignment or fit
of the strategic orientation with respect to the external environment. Levinthal (1997, p935) also states
that “organizational level adaptation leads organizations to modify their existing form in an attempt to
enhance their fitness”. As long as the strategic orientation of the organization fits the one required by
the environment, the organization converges around its current strategic orientation. If for some reason
(e.g. change in the environment) the strategic orientation turns out to be inappropriate then the
organization will change to get a new strategic orientation that would have a better fit.
appropriateness
+
+
fit of SO
-
fit decrease
change in fit
+
perceived fit
+
average
performance
gap
completed changes
felt pressure
to change
+
changes in process
implementation
time
-
+
+
+
new
changes
started
Figure 5: Negative loop describing fit of SO
In the model, this alignment is represented by the variable ‘fit of strategic orientation’(see
figure 5). This is a state variable, because it is the accumulation of the effects of past strategic changes
and the changes in the environment.
t
fit = fit 0 +
fit
t0
and
dfit
= fit incr − fit decr
dt
Fit of strategic orientation signifies the extent to which there is a fit between the actual strategic
orientation of the organization and the one demanded by the environment. The fit gets higher when the
6
organization goes through a successful strategic change (fitincr) and lower when a change in the
environment (e.g. a new law, a better technology, a new product) takes place (fitdecr)1
If the organization goes through change then the fit of its SO will also change. The change in
fit will be effective once the change process is completed. The extent of the effect of change and
whether it leads to improvement in fit depends on the amount of change completed and how
appropriate the new strategy is.
We will first discuss the completed changes and next the appropriateness of a strategy chosen.
3.2 Completed changes
Completed changes are determined by a felt pressure to change, which is in turn determined
by performance as well as inertia. The relationships involved will be discussed below.
3.2.1 Felt pressure to change
The organization starts new changes if there is a felt pressure to change. If the pressure
reaches a value above a particular threshold value, new changes will be initiated. Felt pressure is a
state variable since it signifies the cumulative pressure which builds up over time. The pressure
increase is determined by (a) the perceived fit (the better the fit the lower the pressure) and (b) the
average performance gap (the higher the gap the higher the pressure). Let us discuss each of these two
determinants in turn.
Changes in the external environment constitute a motivation for change. Tushman and
Romanelli emphasize the effect of changes in product class characteristics (demand, technology, users,
and institutional conditions): “As product class conditions change, strategic contingencies faced by the
competing firms also change systematically” (p198). If such a change in strategic contingencies would
generate a misfit between the organization’s strategic orientation and the competitive conditions, this
would be a reason for strategic change. With this reasoning, Tushman and Romanelli propose that the
greater the rate of change in environmental conditions, the greater the frequency of reorientations.2
However, the fact that the fit is low does not mean that the organization will undertake change.
Tushman and Romanelli also emphasize that environmental conditions do not directly cause strategic
reorientation: “While external conditions may be a reason for a strategic reorientation, perception of
inconsistencies is required for its occurence” (p204). Thus, it is the perception of the executives that
guide an organization towards and through a strategic reorientation. This is represented by the variable
‘perceived fit’ in the model. This variable measures the organization-environment fit as perceived by
the executives. If the organization perceives its fit to be lower than the fit required by the environment,
the pressure to change will accumulate. How well the management team perceives the environment is
affected by the organization’s inertia. This will be explained later on.
Along with misfit of the SO, the second factor which leads to an increase in felt pressure to
change is the average performance gap. Tushman and Romanelli consider sustained performance gap
as one of the primary causes of change. Moreover, since organizational change involves risk taking,
the organization needs a motivation to take this risk (Greve, 1998). The organizational risk taking
literature emphasizes the importance of the difference between performance and goals in risk taking
(Greve, 1998; March, 1988). Greve (1998, 1999) ties risk taking to organizational change by showing
that since core changes are risky, organizational performance relative to goals should be an important
determinant of the probability of change. Concordant with this, Tushman and Romanelli state that
realized or anticipated performance pressures constitute the basic driving force for organizational
reorientation. In the model, this is represented as the performance gap. The organization compares its
performance with its goals. The gap between the actual and the desired performance determines the
organization’s felt pressure to change. Sustained performance gap will lead to the accumulation of
pressure to change. While determining the gap, perceived performance rather than actual performance
is used since performance cannot be measured instantly.
1
In the model, the change in the environment and the decrease in fit is produced with a pulse function in the fit
decrease formula.
2
In a frequently changing environment, the organizations should probably have a different attitude towards
change.
7
Now that we have discussed the importance of the average performance gap on initiating
change, we will explain how performance is determined.
3.2.2 Performance
Performance has an important place in the theory as the main driving force behind strategic
change (Tushman and Romanelli, 1985; Meyer et. al., 1995; Greve 1998). It is characterized to be a
consequence of “both appropriate activities with respect to political and economic requirement and of
achieving consistencies in and among organizational activities”. Thus, performance is “tied dually to
fit with the external environment and internal patterns of consistency and interdependence” (p197).
(See figure 6).
Tushman and Romanelli’s explanation of the causes of organizational performance, which has
been dominated by two important theories, has merged these two widespread opinions in organization
theory. With the contingency theory the importance of organization-environment fit as a determinant
of performance has been prevailing, with the view that the appropriate organization form is contingent
on the characteristics required by the environment (White and Hamermesh, 1981, Drazin and van de
Ven, 1985). Thus, to be effective, the organization should achieve a fit between the environmental
requirements and organizational strategy and structure. Congruency theory on the other hand, stresses
the importance of internal consistency between elements of structure (Child, 1972).
Putting the two factors together, in the model, the performance of the organization is
determined by how well the organization fits the environment (fit of SO) and how consistent the
internal operations of the organization (internal consistency) are.
int cons
+ increase
internal consistency
-
change percentage
-
+
?
+
+
fit decrease
appropriateness
+
+
fit of SO
-
change in fit
+
completed changes
felt pressure
to change
+
changes in process
-
performance gap
+
perceived fit
+
implementation
time
performance
+
average
performance
gap
+
+
new
changes
started
Figure 6: causal loop structure of relationships between fit of SO, performance, felt pressure to change
Internal consistency is a state variable since it represents the cumulative achievements of a
company in the way it functions internally. The assumption is that over time internal consistency will
increase as the organization builds up routines after a change process has been completed. However,
internal consistency will decrease when change is going on, and the bigger the amount of change the
organization goes through (i.e. change percentage), the higher the decrease in internal consistency.
8
3.2.3 Inertia and ability to change
An important variable that affects whether a change process will start is ‘ability (willingness)
to change’. This variable is caused by inertia which is an important construct for the theory. Inertia is a
product of the convergent period. It represents “resistance to all but incremental change” (p177).
During the convergent period, the organization strives to get competence, increase internal
coordination, external accountability, and predictability. Generation of these leads to increased
structural and social complexity and interdependence, convergence around a strategic orientation, and
resistance to fundamental change. Thus, as the organization converges around a strategic orientation to
become more and more competent, inertia is generated (see figure 7). In the model, inertia is
represented as a state variable since it accumulates over time as a result of the processes going on in
the convergent period and it decreases when a strategic change takes place. Even though Tushman and
Romanelli explain the processes that go on during the convergent period, these are not explicitly
modeled. We believe that since our main focus is the strategic change, the extensive modeling of all
the processes of the convergent period is not necessary. In order to ensure that inertia is a result of the
convergent period, inertiaincr, which leads to an increase in inertia, is positive only when no strategic
change takes place. As the length of convergent period increases so does inertia. With this
representation, it is assumed that every convergent period of same length would lead to the same
amount of inertia accumulation regardless of the turbulence and success of the convergent period.
Inertia impedes an organization’s ability: (1) “to reassess environmental opportunities and
constraints, and thus to initiate a strategic orientation” (p177), (2) to disrupt relationships and start a
strategic change even when the environmental opportunities are reassessed. The latter effect of inertia
is ‘ability (willingness) to change’. As the organizational inertia increases (decreases), the
organization’s ability to change decreases (increases). Ability to change is given with a decreasing
function of inertia. Ability to change is linked to the threshold which forms an obstacle for the
organization to start a change process. For an organization to change, the felt pressure to change
should be higher than the threshold. If the ability of the organization is high (low) then the threshold is
low (high).
change percentage
change in fit
+
fit
-
+
inertia decrease
+
+ -
completed changes
+
+
perception time
inertia
changes in process
+
-
-
+
-
ability to change
new changes started
+
+
-
inertia increase
threshold
+
perceived fit
felt pressure to change
-
Figure 7: causal loop structure indicating relationships between inertia and changes
The second effect of inertia is related to the executives’ perception of environmentorganization fit, i.e. ‘perceived fit’. Inertial processes affect executive leadership’s decision-making
and reduce the probability of organizational responsiveness to environmental discontinuities. During
the convergent period, the organization elaborates on certain norms, values, and procedures. As the
9
organizational activities become more routinized, the volume and diversity of information processed
decreases. Moreover, more attention is given to information that supports the status quo. Thus, since
inertial forces of convergent periods constraint the perceptions of executives, higher inertia would lead
to higher perception time, i.e. late recognition of organization-environment (mis)fit.
3.3 Appropriateness of strategy
Together with completed changes, the second factor that affects the amount of fit change is the
appropriateness of the strategy chosen3. For an organization to change its strategic orientation, the
organization must first find a new orientation that it wants to achieve. Of course, this can only happen
as a result of searching for information in the environment. In information search for making strategic
decisions, there are two crucial things: the amount of information searched and how critically this
information is processed. According to the threat-rigidity theory (Staw et. al. 1981), “a threat may
result in restriction of information processing, such as narrowing in the field of attention, a
simplification in information codes, or a reduction in the number of channels used” (p502). Thus when
organizations are under threat, then first of all they consult to fewer sources of information. Second,
the new information that comes in is given low value since it cannot be assimilated as easily. Staw et.
al. summarizes the effects of threat on organizations as reliance on poor prior knowledge, a reduction
in communication complexity, and restriction in the information-processing capacity of the
organization.
felt pressure to change
+
+ perceived fit
perceived pressure
+
-
fit
-
information processing capacity
+
quality of information processing
+
change in fit
+
+
accuracy of perception
+
appropriateness
Figure 8: causal loop structure of appropriateness
Since at times of change, the pressure or the threat the organization feels comes from the total
amount of pressure the organization feels upon itself, we have linked the amount of information
searched and how critically this information is processed to the felt pressure to change. When the
organization realizes that it does not fit its environment and that its performance is poor, then the
organization will be under much pressure. Since it will want to get out of the unfavorable situation as
soon as possible, it will look for less information and process the information less critically in order to
find a new strategy4.
In the next sections, we will discuss the model’s dynamics.
4. Model behavior
3
In this paper “appropriateness” is not used to define the fit of the organization to its environment. Here, we are
talking about how appropriate the new strategy chosen by the organization is.
4
Actually, in the feedback loop, the link between perceived pressure and quality of info processing and between
perceived pressure and information processing capacity is not – but rather sometimes – and sometimes +!
10
Following the theory of Tushman and Romanelli, we have postulated some hypotheses to test
our model. For each subsection below, we have formulated an expectation given a certain situation and
run the model for the cases of a young and an old organization. The organizations can be distinguished
as young and old depending on the initial value of their inertia and internal consistency. A young
organization has low inertia and internal consistency whereas an old organization that has not changed
for a long time has a high inertia and internal consistency. For each run, we present two graphs. In the
first graph, one can see 1) new changes started: the amount of new change the organization starts at
each time step, 2) completed changes: the amount of changes that are completed at each time step, 3)
changes in process: the total amount of change processes that are still going on, and 4) fit: the fit of the
strategic orientation. In the second graph, one can see the 1) threshold to change: the minimum amount
of pressure the organization should feel to be able to start change, 2) felt pressure to change: the total
amount of pressure the organization feels to change, 3) fit, 4) internal consistency: the measure of how
well the organization is functioning internally within its current strategic orientation, and 5)
performance: the performance of the organization given the strategic orientation it has and the level of
its internal consistency.
4.1 No change in the environment No change in the organization
Since Tushman and Romanelli say that reorientations will be associated with changes in the
environment and/or sustained low performance, we expect that if no change takes place in the
environment then the organization will not change either5. Moreover following Tushman and
Romanelli’s proposition “holding the external environment constant, the longer and less turbulent the
convergent period, the more effective the organization” (p195), we expect that the longer the
organization does not change, the higher will be the internal consistency and thus the performance.
To test these expectation, we made two runs. The first represents an old organization that has
not changed for a long period, i.e. inertia and the internal consistency are at a high value. As seen in
graph 1, the organization does not change as expected. Since both performance and fit are high enough
the organization has no reason to change.
Then we have made the same run for a young organization or an organization that has just
went through change, i.e. inertia and internal consistency are at a low value. It is expected that such an
organization should converge on the current strategic orientation and do some incremental changes to
increase its internal consistency and performance. In this situation, the positive loop that is responsible
for increasing internal consistency will work. As the convergent period gets longer, i.e. as long as a
change does not take place, the organization will get better and better in its internal functioning.
Increased internal consistency will lead to increased performance. And since the average performance
gap will be closing, the felt pressure to change will not accumulate. Thus the organization will not
change. Figure 10 shows the increasing internal consistency and performance in the absence of
organizational change. Even though at the beginning of the simulation, the felt pressure to change is
higher than the threshold, the organization does not change, because it already fits its environment.
If we would make a run with a young organization with a lower fit of strategic orientation than
the one demanded by the environment then we will see that this organization changes even in the
absence of an environmental change. This behavior does not coincide with the propositions of
Tushman and Romanelli. However, they also say that organizations will change as a result of
performance pressures. A young organization that does not have perfect fit has performance pressures
due to fit problems as well as internal consistency problems. Thus even though such a situation is not
explicitly explained by Tushman and Romanelli, it is very likely that such an organization will start
changing.
5
This is also an equilibrium test. If we start the model at equilibrium, if no disturbance occurs then the system
should stay at equilibrium.
11
n_change_st
1
2
compl_ch
3
ch_in_process
4
fit_div_100
1,000
1,010
1
1 2 3 4
1
n_change_st
2
compl_ch
3
ch_in_process
4
fit_div_100
1 2 3 4
1 2 3 4
1 2 3 4
1 2 3 4
1
2
3
-1,00
4
n_change_st
compl_ch
ch_in_process
fit_div_100
0,990
0
100
200
300
400
500
Time
1,0 3 4 5
3 4 5
3 4 5
3 4 5
1
1
1
3 4 5
3 4 5
3 4 5
3 4 5
3 4 5
3 4 5
3
1
1
1
1
1
1
1
1
0,5 1
2
3
4
5
0,0 2
0
2
2
2
2
50
100
150
200
2
2
2
2
2
2
250
300
350
400
450
500
threshold_to_ch
felt_press_change
fit_div_100
int_consist
performance
Time
Figure 9: runs with no change in environment: old organization
n_change_st
1
2
compl_ch
3
ch_in_process
4
fit_div_100
1,000
1,010
1
1 2 3 4
1
n_change_st
2
compl_ch
3
ch_in_process
4
fit_div_100
1 2 3 4
1 2 3 4
1 2 3 4
1 2 3 4
1
2
3
-1,00
4
n_change_st
compl_ch
ch_in_process
fit_div_100
0,990
0
100
200
300
400
500
Time
1,0 3
3 4 5
3 4 5
3 4 5
3 4 5
3 4 5
3 4 5
3 4 5
3 4 5
3 4 5
3
5
1
1
1
1
1
0,5
2
1
3
1
4
1
1
4 2
0,0 1
0
1 2
50
1
2
100
5
2
2
150
200
2
2
2
2
2
2
250
300
350
400
450
500
Time
Figure 10: runs with no change in environment: young organization
12
threshold_to_ch
felt_press_change
fit_div_100
int_consist
performance
4.2 Change in the environment change in the organization
It is an important proposition of the punctuated equilibrium model of change that
organizational change is triggered by environmental change (Tushman and Romanelli, Meyer et. al.
1995). We will simulate a one time change in the environment by decreasing the fit with a pulse
function. This would mean that a sudden change in the environment such as a new technology or a
new product will be launched to decrease the fit of the organization by 50%. Again we made runs for
young and old organizations. We are expecting both organizations to change because both the fit and
the performance of the organization decreases.
As seen in graph 3, the young organization behaves as expected by going through a change
process. At time 5, a change that takes place in the environment causes the fit of the organization to
decrease further. This causes a decrease in the performance as well. Since this is a young organization,
the inertia is low from the beginning. Thus the organization is apt to the possibility of change with a
low threshold to change. Moreover, the low performance from the beginning coupled with a decreased
fit causes an immediate increase in the felt pressure to change which stimulates a change in the
organization. The change increases the fit and eventually the performance.
1
n_change_st
3,9
2
compl_ch
1,7
3
ch_in_process
42
4
fit_div_100
1,1
4
4
4
2
3
1
4
1
n_change_st
2
compl_ch
3
ch_in_process
4
fit_div_100
2
3
0,0
1 2 3
3 2
1
1
4
1 2 3
n_change_st
compl_ch
ch_in_process
fit_div_100
0,5
0
50
100
150
Time
3
1,0 3
3
4 5
4 5
4
3 5
1
0,5
2
2
5
3
4
2
4
0,0 1 2
1
0
50
1
1
100
5
threshold_to_ch
felt_press_change
fit_div_100
int_consist
performance
2
150
Time
Figure 11: decrease of fit (50%) at time 5: young organization
We have made the same kind of run for the old organization. Contrary to our expectation, the
old organization does not change (figure 12) and the fit remains at 50 rather than returning to its
original value.
1,0 3 4
0,5 1
5
4
4
1
1
2
3 5
3 5
4
1
2
3 5
1
2
3
2
4
5
0,0 2
0
50
100
Time
Figure 12: decrease of fit (50%) at time 5: old organization
13
150
threshold_to_ch
felt_press_change
fit_div_100
int_consist
performance
The reason is as follows. Before the environmental change happens, the pressure to change is
zero. And since an old organization has less ability to change due to high inertia, the threshold that the
felt pressure to change has to overcome is high. The low fit and high performance gap never yield
enough pressure that will surpass the threshold. This behavior looks strange since this would imply
that the old organization would have a sustained gap of 50% in its fit and performance. This ‘invalid’
behavior can be changed by introducing a new piece of structure in the model describing the effects of
a sustained gap. In their model, Tushman and Romanelli explicitly state that a sustained performance
gap is the basic driver for a change process. So far in the model, this is not taken into consideration.
The average performance gap is the only performance component that affects felt pressure to change.
In this particular run, the fact that there is a 0,5 average performance gap has a certain level of
influence on the felt pressure to change. However, the fact that this organization goes on with such a
high performance gap for years and years does not have any influence although sustained low
performance would be a drive for organization to take action and change. We have added this structure
into the model such that on top of the previous two causes, having a performance shortfall for an
extended period of time induces an additional felt pressure to change. When we run the model with
this added structure, we see that the organization changes as expected (see figure 13).
1
n_change_st
28
2
compl_ch
1,8
3
ch_in_process
46
4
fit_div_100
1,1
2 3
4
4
4
1
2
n_change_st
1
2
compl_ch
3
ch_in_process
4
fit_div_100
4
-1
0,0
3
1
1 2 3
1
2 3
4
1
n_change_st
compl_ch
ch_in_process
fit_div_100
0,5
0
50
100
150
Time
1,0 3 4
3
4 5
3
4
2
0,5 1
3 5
1
5
2
1
3
1
2
0,0 2
0
50
100
1
2
4
5
threshold_to_ch
felt_press_change
fit_div_100
int_consist
performance
150
Time
Figure 13: decrease of fit (50%) at time 5: old organization after the addition of the effects of sustained
performance gap on felt pressure to change
In order to check if the improvements we have made in the model change the results of the
runs we have made previously, we have remade those runs. The results did not change. The only
change was that in case of an environmental change, the young organization starts changing a bit
sooner. No graphs have been presented because the old graphs still represent the current behavior.
4.4 Run with a higher inertia level: effect of appropriateness
In the runs of the previous section, for the old organization we had kept the initial inertia at the
level of 0,5. One interesting situation comes about when the initial inertia is set at a higher level, e.g.
0,8 (see figure 14). In this situation, since the ability to change gets low, the threshold to change gets
very high. This means that it takes quite a long time before the organization feels the felt pressure to
change that would exceed the threshold. At the moment the organization feels enough pressure to
14
change, it is also under too much pressure that this affects the information processing capabilities of
the organization negatively, meaning that the organization ends up choosing an inappropriate strategy.
As a result of implementing this new strategic orientation, the fit and the performance gets even lower
and the organization dies
1
n_change_st
43
2
compl_ch
3,8
3
ch_in_process
97
4
fit_div_100
1,0
4
2 3
2
1
2 3
4
2
4
1
n_change_st
2
compl_ch
3
ch_in_process
4
fit_div_100
3
0,0
1 2 3
1
1
200
300
1 2 3
1
4
4
n_change_st
compl_ch
ch_in_process
fit_div_100
###
0
100
400
Time
2
1,0 3 4
4
1
1
4
6
1
0,5
2
5
3 5
3 5
2
4
1 2
1 2
2
2
2
4
1
4
3 5
2
3 5
3
6
0,0 2 6
1 3 5
6
1 3
4
5
1 3
4
4
5
1
5
6
6
6
-0,5
0
50
100
150
200
250
6
300
6
350
threshold_to_ch
felt_press_change
fit_div_100
int_consist
performance
appropriateness
6
400
Time
Figure 14: decrease of fit (50%) at time 5: old organization with an initial inertia of 0,8.
4.4 more changes in the environment more changes in the organization
In order to simulate an environment where sudden changes take place very frequently, we used
a pulse function that causes a 50 decrease every 50 weeks. First we made runs for the young
organization (figure 15) and then for the old (figure 16). Neither of these organizations can survive in
an environment where bursts of change take place at very short intervals. Both of these organizations
will die eventually because of being under too much pressure and thus not being able to formulate
effective strategies.
2
1,0 3
6
3 4
6
3 4 6
3
6
4
5
2
0,0 1
5
2
1
5
2
1
2
1
5
2
4
2
4
0,5
2
2
6
3 4
5
1
6 3 5
1
1
4
2
3
3
1 5
1 3
4
4
5
1
5
6
6
-0,5
0
50
100
150
200
250
300
Time
Figure 15: Young organization in a continuously changing environment
15
6
350
6
400
threshold_to_ch
felt_press_change
fit_div_100
int_consist
performance
appropriateness
2
1,0 3 4
46
4
2
2
2
2
2
2
2
threshold_to_ch
1
0,5 1 5
1
5
2
6
3
3
1
4
4
5
1 3
4
5
1 3
4
5
1 3
5
1 3
5
1
100
performance
5
appropriateness
6
6
200
-0,5
50
int_consist
4
4
6
0
fit_div_100
3
3
5
1
0,0 2 6
felt_press_change
150
6
250
6
300
6
350
6
400
Time
Figure 16: Old organization in a continuously changing environment
A drastic change in the environment every 50 weeks is of course very frequent. The
organization never has the time to relax and build up new routines. It is constantly under pressure. In
these circumstances we can imagine that an organization cannot survive. However, if we would
decrease the number of pulses, in other words, increase the time intervals between consecutive
environmental changes then the organization has time to recover from the previously made changes. In
figure 17a, in the first run environmental change takes place once every year and in the second one,
once every two year. In the second situation, the organization has time to conclude the previous
reorientation before the second environmental change takes place. Thus the fit and performance reach
high values before the organization is faced with a new environment change (see figure 17b). High
performance and fit also lead to low felt pressure to change. However, in the case of environmental
change taking place once every year, since the organization could not finish its change efforts before
the new environmental change took place, the fit and performance could never recover. This caused a
continuous stimulus on the felt pressure to change. And an ever increasing pressure caused the
1,0 3 4
1,01
2
2
2
2
2
3
6
4
46
34
5
6
34
34
6
5
6
34
6
3
45
6
3
5
fit div 100
3
1
1
2
0,5 2
2
0,5 1 5
5
1
5
5
2
3
2
2
4
5
1
0,0
0
1
100
1
200
300
1
400
1
500
0,0 2 6
0
50
2
1
100
150
1
2
200
12
1
1
250
300
350
2
1
6
threshold_to_ch
felt_press_change
fit_div_100
int_consist
performance
appropriateness
400
Time
Tim
organization to choose inferior strategies.
(a)
(b)
Figure 17: (a) fit of the old organization as a result of multiple consecutive change processes. run 1:
environmental change once every year, run 2: environmental change once every two years (b) the old
organization in an environment that changes once every two years.
It is possible that organizations in environments where drastic changes take place frequently
develop other routines, change routines, which would enable them to change more easily. Another
possibility that might be helping such organization survive can be what Hannan and Freeman (1984)
suggest in their theory. According to them most organization fail as a result of their efforts to change.
They propose that those organization that end up in a worse situation as a result of reorientation, fall
back into their old habits and routines. This at least gives the organizations the chance to survive.
16
4.5 Differences young vs. old organizations
According to the literature on organizational change, young and old organizations show different
behavior patterns. Some of these differences are as follows:
•= Young organizations will react to the environmental change faster than the old organizations.
When we compare the reaction time of young and old organizations, we see that after an
environmental change at time 5, the young organization starts changing immediately at time 6
whereas the old organization waits till time 42.
•= High performing organizations have greater losses when changing6 (Greve 1999) and
reorientations in high performing organizations will have negative effects on performance (Meyer
et.al. 1995). If we compare graphs 11 (low performing) and 12 (high performing), we see that the
high performing organization loses performance and has long period of low performance whereas
the young organization always end up with much higher performance than it had started.
Moreover, in frequently changing environments, the young organization survives much longer
than the old organization.
5. Conclusion and discussion
In this paper, we have looked at the theory of Tushman and Romanelli in explaining the way
organizations change. The punctuated equilibrium model says that organizations have long periods of
stability which are altered by radical changes in the face of an environmental change. We have
modeled this strategic change process by using the causal descriptions from Tushman and Romanelli’s
paper. As our starting point, we also took two previously built SD model of this theory, one by
Frechette and Spital and the other by Sastry. Both models were presented and critiqued. Next we
described our model of organizational change, in which we have attempted to eliminate some of the
drawbacks we have found in the previous models. Moreover, two pieces of additional structure were
added. One was related to sustained performance (rather than performance per se) after discovering
that an old organization did not change after a significant decrease in fit. The other relates to the
appropriateness of a chosen strategy, which depends on the quantity and quality of information being
processed. The behavior of the model is largely in line with expectations of the theory and is also able
to simulate conditions under which organizations can be put out of business. For instance when inertia
level is too high, but also in cases in which the organization is under too much pressure. The latter
affects their effectiveness in the way they search for information and the way they process
information. This might lead the organization to choose an inferior strategy and in certain cases puts it
out of business.
Naturally our model is still far from being complete. Several things discussed in Tushmann
and Romanelli’s theory and in related papers need to be added to future versions of the model. One of
these concerns the level of desired performance. At the moment, in the model, desired performance is
held fixed. However, Tushman and Romanelli explicitly say that change in power distribution may
lead to changes in the performance aspiration levels in the organization. The theories of risk taking
also emphasize the importance of changing aspiration levels. March (1988) summarizes these ideas as
follows: “Understanding of variable risk preferences will require not only making preferences for risk
dependent on a target, but also making the target itself a function of experience” (p232). Greve (1998)
mentions social and historical aspiration levels. Social aspiration levels are determined by the
performance of similar ones whereas historical aspiration levels are determined by the performance
history of the firm itself. Since Tushman and Romanelli restrict themselves to the level of one
organization without paying attention to the effect of other organization’s behavior, new versions of
the model will focus on historical performance. Thus, in the model the desired performance will get
updated during the simulation. Initially it is set equal to 1, which is the maximum performance level
an organization can reach. March states that the slower adaptation of goals makes action sensitive to
6
In our case, an old organization is an organization that has high performance (100 fit, 1 internal consistency)
and a young organization means an organization with low performance (100 fit, but low internal consistency)
17
performance history. Thus, the goal is not updated as a result of instantaneous performance gaps but
rather the sustained average performance gap. If there is a sustained performance gap, the organization
lowers its required performance. This structure represents the ‘drifting goals’ archetype in system
dynamics. When a persistent performance gap exists the organization takes corrective action
(increasing pressures for change) as well as lowering its goals. Thus, in the following versions of the
model, desired performance will be variable.
A second improvement concerns the success of change processes. An assumption in Sastry’s
model seems to be that a change attempt always results with a success. Tushman and Romanelli’s
theory also does not pay much attention to the success issue. Given the vast number of unsuccessful
change attempts in the industry, it would be logical to relax such an assumption and look at the
behavior when a change attempt does not result in full success.
18
6. References
Child, J. 1972. Organizational Structure, Environment, and Performance: The role of strategic choice.
Sociology 6: 2-22.
Drazin, R. and A. H. van de Ven. 1985. Alternative Forms of Fit in Contingency Theory.
Administrative Science Quarterly 30: 514-539.
Frechette, H. M., F. Spital. 1991. A Model of Organizational Change. Proceedings of the 9th ISDC:
209-218.
Frechette, H. M., F. Spital. 1992. Organizational Design and the Dynamics of Performance over the
Long Term. Proceedings of the 10th ISDC: 183-189.
Greve, H. R. 1998. Performance, Aspirations, and Risky Organizational Change. Administrative
Science Quarterly 44: 58-86
Gersick, C. J. G. 1991. Revolutionary Change Theories: A Multilevel Exploration of the Punctuated
Equilibrium Paradigm. Academy of Management Journal 16(1): 10-36.
Greve, H. R. 1999. The Effect of Core Change on Performance: Inertia and Regression Towards the
Mean. Administrative Science Quarterly 44: 590-614.
Hannan, M. and J. Freeman. 1984. Structural Inertia and Organizational Change. American
Sociological Review 82: 929-964.
Levinthal, D. A. 1997. Adaptation on Rugged Landscapes. Management Science 43(7): 934-950
March, J. G. 1988. Variable Risk Preferences and Adaptive Aspirations. Journal of Economic
Behavior and Organizations 9: 5-24.
Meyer, A. D., J. B. Goes, and G. R. Brooks. 1995. Organizations Reacting to Hyperturbulence. In
Huber, G. P. and A. H. van de Ven (Eds.) Longitudinal Field Research Methods Studying Processes of
Organizational Change: 299-349. Informs, Sage Publ. Inc. California.
Sastry, M. A. 1997. Problems and Paradoxes in a Model of Punctuated Organizational Change.
Administrative Science Quarterly 42: 237-275.
Staw, B. M., L. E. Sandelands, and J. E. Dutton. 1981. Threat-Rigidity Effects in Organizational
Behavior: A Multilevel Analysis. Administrative Science Quarterly 26: 501-524.
Tushman, M. and E. Romanelli. 1985. Organizational Evolution: A Metamorphosis Model of
Convergence and Reorientation. In L. L. Cummings and B. M. Staw (Eds.). Research in
Organizational Behavior 7: 171-222. Greenwich, CT: JAI Press.
White, R. E. and R. G. Hamermesh. 1981. Toward a Model of Business Unit Performance: An
Integrative Approach. Academy of Management Review 6(2): 213-223.
19